Prediction

sorenbirk skrev:
> I need to predict the future value of a signal. I dont know anything about
> the signal other than it looks like bandpass filtered white noise.

This is the killer. Prediction is based on some model of the
signal, some sort of structure that relates sample n+d (d>0)
to sample n.

> I need
> to predict the value say 10 samples ahead (I guess this is called a
> ten-step predictor???). Does anyone has a good suggestion on how to do
> this? Linear prediction? Adaptive filtering or system identification
> techniques?

While there exist methods to do this by linear prediction,
your lack of knowledge about the signal has already excluded
these from consideration. LP works, at least to some extent,
provided you can justify the AR(p) model. If you can't justify
why AR(p) fits, use a random number generator instead.

Rune